Readout Efficiency Tests

For the Key Project we are interested in temporal sampling of the PSF. So here we did some efficiency tests of the various combinations of Integration time, Coadds, and Cubes to see which mode to use to jointly optimize temporal sampling and efficiency. First let's explain the different options in ways to save Coadds and Cubes.

Clio Data Sequence Options:

Between Coadds, Nimgs, and Cubes, there are a few options for temporal sampling and resultant image size for how to take a sequence of Clio data. If you are using Cubes, then the Coadds are not added but are saved as individual frames within a 3-dimensional data cube. The Cube will be saved as a single .fit file and so it will have only 1 FITS header. Thus, while the temporal sampling of the data is higher, the temporal sampling of the FITS header (not listed here) is per .fit written. Also, the FITS header will state the # of Coadds only if they are added, so if you are saving in Cubes, asking for "10 coadds saved in cubes" gives you 1 coadd in 10 frames in a data cube.

The Table explains it with an example:
Nimgs Integration Time [ms] Coadds Cubes What Will the FITS Header Say for # of Coadds Result FITS dimensions (Example for Full Frame) Total Integration Time [sec] = Nimg x IntTime x Coadds Temporal Sampling of Images [sec]
10 1000 1 No 1 10 2-d FITS files, no coadds 1024 x 512 10 1
10 1000 10 No 10 10 2-d FITS files, each with 10 frames coadded (the total duration is 10x longer than the above line) 1024 x 512 100 10
10 1000 1 Yes 1 10 "3-d" FITS files, with the coadds saved in cubes except that since there is only 1 coadd, the 3rd dimension has nothing in it 1024 x 512 x 1 10 1
10 1000 10 Yes 1 10 3-d FITS files, each with 10 frames in a 3-d Cube 1024 x 512 x 10 100 1




And here are the readout efficiency tests:

(Also a quick note about taking data at a very high temporal sampling. To save computer processing speed, you may want to stop the Transfer Files program (hit the X in upper left corner of window) during the data taking. Restart it when resuming normal observations by going to Restart Clio Components -> FileBackup)


Strip Mode Tests:

Strip Mode
Integration time [ms] # Coadds Cubes or 2-d FITS? # Frames Images ID Total Clock Time Elapsed [min:sec] / [sec] Total Integration [sec] % Efficiency = 100 x Total Integration / Clock Time Elapsed Temporal Sampling of Images [sec]
164 40 Cubes 15 n140410 test 1-15 3:39.5 / 219.5 98.4 44.8 % 0.164
164 40 2-d FITS 15 n140410 test 16-30 3:33.3 / 213.3 98.4 46.1 % 6.56
164 1 2-d FITS 600 n140410 test 31-630 6:30.0 / 390.0 98.4 25.2 % 0.164
164 6 Cubes 100 n140410 test 631-730 4:10.8 / 250.8 98.4 39.2 % 0.164



Stamp Mode Tests:

Stamp Mode
Integration time [ms] # Coadds Cubes or 2-d FITS? # Frames Images ID Total Clock Time Elapsed [min:sec] / [sec] Total Integration [sec] % Efficiency = 100 x Total Integration / Clock Time Elapsed Temporal Sampling of Images [sec]
1000 10 2-d FITS 10 n140406 AlphaCen 1-10 2:14.2 / 134.2 100 74.5 % 10
10000 1 2-d FITS 10 n140406 AlphaCen 11-20 1:47.8 / 107.8 100 92.8 % 10
1000 1 2-d FITS 100 n140406 AlphaCen 21-120 2:50.0 / 170.0 100 58.8 % 1
43 100 2-d FITS 10 n140406 AlphaCen 121-130 1:47.4 / 107.4 43 40.0 % 4.3
43 10 2-d FITS 100 n140406 AlphaCen 131-230 2:08.6 / 128.6 43 33.4 % 0.43
43 25 2-d FITS 40 n140407 AlphaCenTest 1-40 1:57.4 / 117.4 43 36.6 % 1.075





-Katie Morzinski, 2014/07/17, tested 2014/04/06-10